Swotti: Claims of Semantic Review Aggregation are Highly Exaggerated

Nothing excites me more than hearing about semantic anything. I'm lazy. I have too many things to read already. And I am tired of trying to sift through tons of review sites where reviews are often incomplete, non-existant, or just plain irrelevant. I was thrilled to hear about Swotti, a new beta site that claims it will do all that sifting work for me, leaving me with a crisp, clean snapshot review of anything I'm looking for.

The reality is that Swotti fall far short of a true semantic application, and nowhere is it more evident than on a popular item. I can't even remember what the first few items I searched for were, because no sooner did I hit enter than I was told there wasn't enough information for my query. Assuming that the popular iPhone would have to net enough information, I went with that.

Swotti is less a semantic search than a keyword-optimized search. If you are handy with SEO, odds are you can get your site aggregated into Swotti's app, no matter how dubious the relevance. Listed in the first 15 results are one result for the iPod Touch, two “free” articles listed by different authors with identical text for iPhone cases, one for a display protector, and a forum topic asking for an iPhone template for Open Source project Joomla. In other words, 1/3 of the first 15 results aren't even relevant to my search. Since the splog fodder articles contain iPhone (spelled Iphone, no less), best, and stylish, those go into the “positive” review aggregation. The same goes for the forum topic (fast, clean and friendly along with iPhone) and the screen protector (love and “highly recommend”).

Swotti is no more a semantic search for reviews at this point than Google itself is, and I'm surprised at the positive reviews so far. If I'm going to have to add boolean search terms to my searches there, it's no more helpful to me than a regular web search, and actually, adds more steps to the process.